High volume distributed production systems with many points of modification are subject to frequent defects due to their complexity and the complexity of managing these solutions. Previous solutions have included attempts by companies to provide self-healing systems and redundant (“failover”) systems. However, these existing systems rely upon the failover system itself to be either completely unified under the company's platform, or at the very least to be provided or hosted by the company. Further, these existing systems are not expandable to incorporate new error scenarios, making them unadaptable. They also cannot capture human expert solutions, and are extremely complicated to install and use.
A solution is needed in which an automatic defect resolution system operates independently of the environment and systems which it administers and can be turned off and on without adversely impacting production running. An automatic defect resolution system that operates by parallel log monitoring of existing and new state systems would resolve many of these problems.
Ideally, such systems would not require advanced technical knowledge to install, operate or maintain, making them extremely simple to install and use. More elegantly designed systems would avoid problems associated with currently existing high overhead/high architecture solutions, which have limited expandability into unknown errors. Human expert solutions could be captured and expanded to capture any future error scenarios. Further, interval based transaction tolerances could be used to mathematically guarantee performance within intervals. Most of all, an ideal solution would eliminate delays associated with promoting changes to the production system.
Accordingly, such an automated parallel deployment system would resolve many of the shortcomings currently present in existing defect solutions.